Greetings! First post here...

Looking for help on a 2016 L405 3.0L TDv6 single turbo.

For about the first 2 minutes after starting the idle keeps hunting/surging. Inspecting things under the bonnet I can hear fluid gushing sound from the Pressure Control Valve (PCV) fuel return line. Touching that return line I can feel it swell up a bit by pressure. Twice every second the PCV is being commanded to release almost all pressure from the high pressure fuel rail.

Have confirmed both by observing the PIDs on a scan tool and by putting a scope on the PCV control wires. The ECU starts out with about 38% duty cycle on the PCV and keeps reducing it rapidly until it hits around 9 ~ 10%. At which point the engine begins to stall and the ECU immediately increases the duty cycle back to 40%. The engine recovers and the cycle repeats again. Happens almost twice every second.

This happens no matter if the engine is cold or is at operating temp. This hunting/surging starts happening from the very moment the engine is started.

If I attempts to drive the car when this surging is happening the whole car turns into a bucking bronco. The ECU continues to ramp down and then step back up the duty cycle of the PCV in a saw tooth manner even with light application of throttle.

During all this the duty cycle of the metering valve on the CP4 pump is being maintained at around 30%. The rail pressure sensor correctly reports the pressure in the rail. I can even see what I think are the spikes of the injectors opening when looking at the output of the pressure sensor on the oscilloscope. Not sure if it is the sensor that is picking the pressure change spikes or its just getting induced into the probes from the injector wires. When the overall pressure in rail drops significantly, those spikes disappear and then show up again when the pressure increases above a certain threshold.

Around the 2 minute mark after starting the engine the idle smooths out. The ECU appears to go into a different operating mode. The duty cycle of the PCV changes to ~37% and stays steady. The fuel metering valve changes to 40%. Once this happens I can then drive the car and all appears to be ok. Power seems fine. Overall mpg is low and the exhaust smells like that of a diesel truck from the 70s.

At this point with engine running smoothly if I shut it off and start is up again right away then it goes back into this hunting/surging idle mode. Give it around 2 minutes and it smooths out again.

Here's the kicker. If I loosen the PCV just the right amount and allow just a small amount of pressure to leak from the high pressure side of the fuel rail to the low pressure return line the the engine does not have this surging idle problem. The ECU starts out the PCV at around 37% and very gradually brings it down to ~30%. The metering valve is kept steady at around 32% duty cycle. The engine idles smoothly. Then around the 2 minute mark, right about the same time when the car would go from bucking idle to smooth idle when the PCV is torqued down to spec, the duty cycle of of both the PCV and metering valve is switch to ~ 37%. Engine continues to idle smoothly.

The only reason I even know that I can get it to idle smoothly with the PCV not torqued down to spec is because the car came to me in this state. Whoever was working on it before me somehow discovered it.

The issue with leaving the PCV not fully torqued down besides the obvious safety concerns is that the fuel rail tops out at 56 Mpa. When the commanded pressure and actual pressure start to drift apart, the ECU shuts the engine off and throws a P0087.

Speaking of codes, with the PCV properly torqued down the ECU throws absolutely no codes. None. So far have consumed about 5 gallons of diesel in driving and idling and there are no codes what-so-ever.

Now the history of the car that I've managed to put together based on some paperwork that was in the car at the time of auction and what is installed on the car. The fueling system on it was replaced. As best as I can tell everything except for the tank and the lines from the tank to the fuel filter was replaced. CP4 appears to be remanufactured unit. The fuel rails and injectors are new and genuine Bosch. The reason for the repair stated on the paperwork is Vandalism - Sand introduced to fuel tank. I took lift pump out of the tank. Found a couple of dead insects floating around but no sand. Took apart the lift pump as much as I could and its screens were clean. No tell tale signs of sand in there.

Stuff I have checked:

1. Powers and grounds. Wiring appears to be fine to/from the ECU.

2. Visual inspection other than signs of frustration from whoever was working on it before, I dont see anything wrong in there.

3. Replaced the PCV with a used one I had sitting on the shelf. No change. Bought a new PCV and installed it on the left side rail I had sitting on the shelf. Installed it in the car. No change. With all 3 PCVs I can reproduce the same behaviour of smoothing out the idle if I intentionally leave the PCV loosened just the right amount.

4. Replaced the metering valve with a genuine Bosch. No change. The one on there was a Chinese clone.

5. Replaced the rail pressure sensor no change. Though with the pressure sensor disconnected the engine starts bucking/surging almost 3 to 4 times every second instead of twice every second and a bit more violently.

6. Checked low pressure side. It is a little low reading at ~ 6 psi. I plumbed in a pump in series which boosted the pressure on the low side to 12 psi. No change.

7. Checked the internal Pressure Relief Valve (aka overflow valve) on the CP4 by removing it. Its got a very small amount of rust pitting on it but appears to be working.

8. Checked the IQA codes stored in the ECU. They are correct.

9. Have run the "Powertrain set-up - Diesel fueling adaption re-initialization" and "Powertrain - control module adaption clear" service functions, no change.

10. Have left the battery disconnected for a week. No change. Been working on this lump for 3 months now (weekends only.)

Things I have not done yet:

1. Checked the timing of the CP4 pump. Mind you after the first 2 minutes and when the idle smooths out I can drive the car no problem.

2. Replaced the CP4 pump. Hmm... Dont want to do that if I dont have to. All the lines and rails I have taken apart so far have had no signs that I've found of the internals grinding away.

So if you have made it this far, thank you! Would love to hear any suggestions you may have. I have some videos of the scope captures. Will post them once I upload them.

Shahid

Here's the first video.

Yellow trace: Pressure Control Valve (PCV)
Blue trace: Fuel Metering Valve
Magenta trace & Fluke meter: Fuel Rail Pressure sensor

The Foxwell scan tool has PCV duty cycle, Metering valve duty cycle, and Fuel Rail Pressure sensor voltage on it.

Both valves are ground switched. When the PWM signal drops low is when the valve is energized. The PCV closes when energized. The Metering Valve opens when energized.

Immediately after starting it goes into this mode where it starts aggressively reducing the duty cycle of the PCV until fuel pressure can no longer be sustained. As soon as the pressure drops low enough that the PCM stops firing the injectors, it increases the duty cycle on the PCV and the pressure in the rail starts to recover. When the pressure sensor reaches about ~ 1.0V the whole cycle starts again.

At 2:06 in the video you'll see that the PCM apparently changes modes and the engine smooths out. There is a change in the Metering Valve duty cycle from ~31% to 41%. PCV gets commanded with a steady duty cycle instead of the rapidly varying one. And the Rail Pressure stabilizes at ~ 0.69V.

This happens every time. Regardless of the engine being cold or up to temperature. Idle fluctuates like this for about 2 minutes and then stabilizes.

Exhaust is absolutely noxious. Cannot stand near the car or sit in it with the windows down.

Video with time base slowed down to show the ramping up of fuel rail pressure and its sudden drop. This is the result of how the PCM is commanding the PCV.

The spikes you see on the magenta trace are the injectors firing. The FRP sensor wiring runs in the same loom as the injector wires and the injectors being fired is getting picked up by the scope because of cross-talk.

You can see the injectors stop firing when the pressure drops too low.

And finally this is how it behaves if I intentionally loosen the PCV and let a small amount of pressure bleed off all the time from the high pressure side to the low pressure side.

Exhaust is still really bad with it running like this.

Another test I have done is to supply the FRP sensor with 4.6V from a bench power supply instead of the 5V that the PCM supplies it. With the intended goal being to proportionally lower the output voltage hence making the PCM think that the rail pressure is lower than it really is.

By doing this and making sure that the sensor output never exceeds 0.8V during the first few revolutions of the engine I can get the engine to run smoothly. However that only works on a cold engine. When the temps around the FRP sensor start to go up, it no longer functions reliably at 4.6v and shuts off.

At this point I am leaning towards the high pressure fuel pump being bad. Its a remanufactured pump. I think its putting out more pressure than the PCM expects. Causing it to panic and going into some sort of an unstable feedback control loop.

I have replaced the metering valve with a couple of other ones from different cars. All behave the same way.

Wow! This is a bit different to a "My car has an annoying squeak" post.
I wonder if HPFP timing would have any effect. There's a debate about the pump pulsing as the injectors open to maintain pressure so, if the timing was wrong it would build excess pressure during the injector closed phase. But why it would fix itself after 2 minutes.......?
Hope you get to the bottom of it. Please keep us informed.

Hi ss10, not sure if this helps but what i noticed in your write up is the very high pressure of 56 Mpa.
Please see attached picture where it states 26.5 Mpa at idle speed for my 2020 SDV6 275 hp.
Is this 56 Mpa not excessive?

Spridget wrote:

I wonder if HPFP timing would have any effect.

I cant imagine it would have a problem like this if the pump was out of time. Whatever happens that causes the PCM to throw a fit happens in the first few hundred milliseconds after starting.

But I have gone down the pump timing rabbit hole as well. On a clean scope capture of just the rail pressure I can see each compression stroke of the pumping elements. Pointed by the red arrow below. That coincides with the crosstalk from the injectors opening. The green arrow. You can see this to some extent in the second video above as well.

janpel wrote:

Is this 56 Mpa not excessive?

ss10 wrote:

The Metering Valve opens when energized.

Correction. The Metering Valve is closed when energized. My understanding of how it worked was incorrect.

During my troubleshooting I accidentally left the metering valve disconnected. Was surprised to find that the engine behaved exactly the same way.

Energized the metering valve from an external power supply and the car did not start. Fuel was completely cut-off.

Took the metering valves I have apart. The valve barrel is made with very high precision. In the picture below the left most barrel is from a counterfeit valve which was on the car when I got it. The middle one is a used one off of a wrecked but running car with TDv6. Right is a new one from the valve for a TDv8.

Each barrel has three openings. The picture is that of the smallest opening on all 3. The counterfeit one is junk. I have two of them, they both are the same.




Same picture but with a scale. Each graduation is 1mm.



So I believe my issue is the metering valve. Its an over-fueling issue.

Have also looked at the spare parts for the CP4 pump. Doesn't appear to be any differences in the pumping elements from one pump to the other except for the orientation of out nipple.

Would someone with a TDv6 be willing to do a test for me. If you disconnect the metering valve and start the car it should start idling the same way mine does. This will throw a code and check engine light so you would have to clear that. Thanks in advance if you take this on.

Shahid

Capture of actual and commanded rail pressures through SDD.

https://www.fullfatrr.com/gallery/displayi...fullsize=1